Abstract: The rapid emergence of new legal terminology – driven by technological innovation, digital governance, and cross-border regulatory expansion – has intensified challenges in multilingual legal communication. Traditional lexicographic and doctrinal approaches struggle to capture the dynamic evolution of terms in areas such as cyber law, fintech regulation, and AI governance, particularly where meanings shift across jurisdictions and languages. This study addresses the problem by modelling the emergence, transformation, and conceptual grounding of modern legal terminology through a combined computational–ontological framework. The research is theoretically anchored in computational linguistics, legal semiotics, and ontology engineering, positioning legal terms as both linguistic signs and nodes within structured conceptual systems. To operationalise this perspective, the study integrates neural language models (BERT, GPT, RoBERTa) with domain-specific ontologies to detect emergent terminology, trace semantic change, and align new terms with established legal concepts. The methodology employs multilingual corpus construction, corpus-driven term extraction, contextual semantic embedding analysis, clustering techniques, and ontology alignment using OWL and SKOS-based frameworks. Semantic drift and conceptual novelty are quantified through embedding distance metrics and expert-validated mappings. The results are expected to contribute to the harmonisation of legal terminology across languages, reduce ambiguity in legal translation, and support more consistent legislative drafting. By combining the predictive and discovery capacities of neural models with the formal structure of ontologies, the proposed framework offers a scalable approach for monitoring terminological evolution and enhancing semantic interoperability in contemporary legal systems.

Keywords: Neural language models, legal terminology emergence, ontological frameworks, semantic interoperability; legal semiotics; BERT, GPT, RoBERTa, Corpus-based term extraction, semantic embeddings, ontology alignment.

Download: | DOI: 10.17148/IMRJR.2025.021106